Respected Sir,
I'm trying to validate my model for tides and currents. The model is giving excellent results for tides, but it is heavily underestimating the currents. The current velocity around Goa region is of the order of 0.3 m/s, but my model is predicting around 0.15 m/s. I also tried using the Blumberg Kantha radiation scheme at the open boundaries to boost up the kinetic energy, but it didn't help. Can you give me some guidance as to which parameters should I change so that the current velocities are increased? Also, the maximum water depth in my domain is around 60 metres. Can you suggest me a decent relaxation time for my TLAG file? I have used 200 seconds in deep water, and gradually increased it to 1800s near the coast, as suggested in the wiki. Should I reduce it, as the water depth is only 60m, or should I increase it?
Validating the model

 MOHID Beginner
 Posts: 16
 Joined: Tue Jun 24, 2014 07:19

 MOHID Beginner
 Posts: 16
 Joined: Tue Jun 24, 2014 07:19
Re: Validating the model
I got some very useful tips from Luis Sir. He wrote:
When validating the model here is a list of things you need to take into account:
1) Is the comparison for currents made based on depthintegrated model results? Did you run the model in 2D or 3D?
2) How were the currents measured? Is this information for a point in the water column, or depth integrated data from an ADCP?
3) What’s the horizontal resolution of your model? Is the bathymetry in the grid cell where the comparison is made really representative of the real bathymetry? Was the currentmeter/ADCP placed in a channel? If so, is this channel correctly discretized by your computational grid?
4) Are you forcing the model only with tide? What about wind? Fresh water discharges? Density gradients? Waves? All this factors influence currents and can account for the “remaining 0.15m/s” that you say you’re missing in the model results.
The Blumberg Kantha radiation scheme will help you solve problems with “numerical noise” reaching the open boundary and handling it. If the model results are excellent for water elevations this means you’re on the right track (and have no problems with this kind of “noise”). You just need to add more physical forcing to the model and possibly increase horizontal and/or vertical resolution.
Regards,
Luis
When validating the model here is a list of things you need to take into account:
1) Is the comparison for currents made based on depthintegrated model results? Did you run the model in 2D or 3D?
2) How were the currents measured? Is this information for a point in the water column, or depth integrated data from an ADCP?
3) What’s the horizontal resolution of your model? Is the bathymetry in the grid cell where the comparison is made really representative of the real bathymetry? Was the currentmeter/ADCP placed in a channel? If so, is this channel correctly discretized by your computational grid?
4) Are you forcing the model only with tide? What about wind? Fresh water discharges? Density gradients? Waves? All this factors influence currents and can account for the “remaining 0.15m/s” that you say you’re missing in the model results.
The Blumberg Kantha radiation scheme will help you solve problems with “numerical noise” reaching the open boundary and handling it. If the model results are excellent for water elevations this means you’re on the right track (and have no problems with this kind of “noise”). You just need to add more physical forcing to the model and possibly increase horizontal and/or vertical resolution.
Regards,
Luis
Who is online
Users browsing this forum: No registered users and 1 guest